Recent machine learning-driven investigations into high entropy alloys: a comprehensive review

Abstract

The exploration of high entropy alloys (HEAs) primarily relies on trial-and-error experiments and multiscale modelling, which are time-consuming and resource-intensive. Recently, machine learning (ML) has emerged as a powerful tool for studying HEAs by training models and identifying key features. This review summarizes the recent progress in applying ML methods to HEAs, with a particular emphasis on properties relevant to structural materials. First, we outline the general process of utilizing ML methods in materials science. We then systematically review recent applications of ML in design of HEAs, focusing specifically on phase formation, structural properties, and constructing interatomic potentials. Finally, we highlight significant challenges and future research directions for ML methods in HEAs, offering insights into ongoing exploration and development.